寻求WebRTC Native Android开发完整文档与后续操作指引

一、必要权限配置 #

• 在AndroidManifest.xml中声明静态权限：

  <uses-permission android:name="android.permission.RECORD_AUDIO" />
  <uses-permission android:name="android.permission.INTERNET" />
  <uses-permission android:name="android.permission.ACCESS_NETWORK_STATE" />
  <uses-permission android:name="android.permission.MODIFY_AUDIO_SETTINGS" />
  

• Android 6.0+需动态申请录音权限：

  private const val RECORD_AUDIO_PERMISSION_CODE = 1001
  
  // 在Activity/Fragment中执行申请逻辑
  if (ContextCompat.checkSelfPermission(this, Manifest.permission.RECORD_AUDIO) != PackageManager.PERMISSION_GRANTED) {
      ActivityCompat.requestPermissions(this, arrayOf(Manifest.permission.RECORD_AUDIO), RECORD_AUDIO_PERMISSION_CODE)
  }
  

二、核心组件初始化 #

1. 初始化PeerConnectionFactory #

这是WebRTC的核心工厂类，需先完成全局初始化：

// 建议在Application类中执行全局初始化
val initOptions = PeerConnectionFactory.InitializationOptions.builder(applicationContext)
    .setEnableInternalTracer(true)
    .createInitializationOptions()
PeerConnectionFactory.initialize(initOptions)

// 创建工厂实例
val factoryOptions = PeerConnectionFactory.Options()
val peerConnectionFactory = PeerConnectionFactory.builder()
    .setOptions(factoryOptions)
    .setAudioEncoderFactory(DefaultAudioEncoderFactory(
        EglBase.create().eglBaseContext,
        true,
        true
    ))
    .setAudioDecoderFactory(DefaultAudioDecoderFactory(EglBase.create().eglBaseContext))
    .createPeerConnectionFactory()

2. 创建本地音频轨道 #

用于捕获设备麦克风的音频数据：

val audioConstraints = MediaConstraints()
val audioSource = peerConnectionFactory.createAudioSource(audioConstraints)
val localAudioTrack = peerConnectionFactory.createAudioTrack("local-audio-track", audioSource)

三、信令交互（核心逻辑） #

WebRTC依赖信令服务器交换SDP（会话描述）和ICE（网络候选），需实现以下关键流程：

• 本地生成Offer/Answer后，通过信令发送给远端
• 接收远端SDP，设置为PeerConnection的远程描述
• 收集本地ICE候选，发送给远端
• 接收远端ICE候选，添加到PeerConnection

示例代码片段：

// 配置ICE服务器（使用谷歌公开STUN服务器）
val iceServers = listOf(PeerConnection.IceServer.builder("stun:stun.l.google.com:19302").createIceServer())
val rtcConfig = PeerConnection.RTCConfiguration(iceServers)

// 创建PeerConnection并设置观察者
val peerConnection = peerConnectionFactory.createPeerConnection(rtcConfig, object : PeerConnection.Observer {
    override fun onIceCandidate(candidate: IceCandidate?) {
        // 将ICE候选通过信令发送给远端
        candidate?.let { sendIceCandidateToRemote(it) }
    }

    override fun onAddStream(stream: MediaStream?) {
        // 处理远端音频流，实现播放逻辑
        stream?.audioTracks?.firstOrNull()?.let { remoteAudioTrack ->
            val audioSink = object : AudioTrack.AudioSink {
                override fun onData(audioData: ByteBuffer?, bitsPerSample: Int, sampleRate: Int, numberOfChannels: Int, numberOfFrames: Int) {
                    // 此处可将音频数据传递给Android音频系统播放
                }
            }
            remoteAudioTrack.addSink(audioSink)
        }
    }

    // 实现其他必要的Observer方法
    override fun onIceConnectionChange(newState: PeerConnection.IceConnectionState?) {}
    override fun onSignalingChange(newState: PeerConnection.SignalingState?) {}
    override fun onIceGatheringChange(newState: PeerConnection.IceGatheringState?) {}
    override fun onRemoveStream(stream: MediaStream?) {}
    override fun onDataChannel(dataChannel: DataChannel?) {}
    override fun onRenegotiationNeeded() {}
})

// 添加本地音频轨道到PeerConnection
val localMediaStream = peerConnectionFactory.createLocalMediaStream("local-stream")
localMediaStream.addTrack(localAudioTrack)
peerConnection.addStream(localMediaStream)

// 创建Offer并发送给远端
peerConnection.createOffer(object : SdpObserver {
    override fun onCreateSuccess(desc: SessionDescription?) {
        peerConnection.setLocalDescription(this, desc)
        // 将本地SDP通过信令发送给远端
        desc?.let { sendSdpToRemote(it) }
    }

    override fun onSetSuccess() {}
    override fun onCreateFailure(error: String?) {}
    override fun onSetFailure(error: String?) {}
}, MediaConstraints())

四、资源释放 #

通话结束后，需及时释放WebRTC资源避免内存泄漏：

localAudioTrack.dispose()
peerConnection.close()
peerConnectionFactory.dispose()

五、实用参考 #

• 参考WebRTC官方示例App（AppRTCMobile）：包含完整的语音通话逻辑，涵盖信令交互、权限处理、音频渲染等核心流程
• 官方Native文档中PeerConnection生命周期、音频轨道配置的说明，可帮助理解核心组件的工作机制

内容的提问来源于stack exchange，提问作者Md Ahsan Kabir